June 20, 1959 And.
ST.\BILITY
AND S Y N T H E S I S OF P H E N Y L B O R O N I C
Calcd. for C21H140d: C, 50.24; H, 4.49.
C, 79.50; H, 4.49.
Found:
Hydrolytic Cleavage of I and VIII.-2,2'-Tolandihoronic acid (0.507 g.) and 0.4 nil. of w:iter were sealed in a Pyrex tube and heated for 29 hours a t 200' in a n autoclave. Compound LTIII(0.331 g . ) and 0.4nil. of water were treated similarly. Both reactions yielded yellow oils which were ether soluble and a white solid (boric acid) which was insoluble in ether. T h e ether solutions were washed with s( diutn hydroxide solution and evaporated. T h e product from VI11 solidified (m.p. 25-45'), and the infrared spectrum was essentially the same as t h a t of a n authentic sample of desoxyhenzoin. T h e residue from the reaction of I failed to crystallize. Both the product from VI11 and I, however, yielded a 2,4-dinitrophenylhydrazone derivative which melted a t 201-202'. Desoxybenzoin 2,4-dinitrophenylhydrazone melted at 200-202" and the melting point was not depressed on admixture with the dinitrophenylhydrazone derivatives of the products of I and VIII. Test Series of Conversion of I to VII1.-Unless otherwise noted, the solutions contained in each case approximately 0.3 g. (accurately weighed) of 2,2'-tolandiboronic acid, 15 ml. of water, 15 ml. of 95% ethanol and the added reagents (sodium hydroxide, tartaric acid, etc.) as indicated in Table I. They were heated at reflux for 3 hours and then cooled, acidified with hydrochloric acid and stripped of most of the solvent in vacuo. The resulting precipitates were collected, washed with sodium bicarbonate and with water and tlried at 55'. Infrared spectra were taken in potas-
[CONTRIBUTION FROM
THR
3017
ACIDS
sium bromide pellets which were 2% sample by weight. The significant features of the spectra of I, V I I I , and one of the mixtures of I and VI11 (from the sodium acetate reaction) are shown in Fig. 3. Increasing percentages of VIII were marked by increasing intensities of the bands at 6.1, 11.1, 12.1 and 12.9 p and decreasing intensities of the bands at 6.25, 9.55 and 13.2 p . ( I and VI11 absorbed strongly in the B-0 region, 7.25 p , and the 0-H region, 2.95 p for I and 2.90 p for VIII. Neither absorbed in the region for anhydrides of boronic acids, 14-15 r.21) It was found from known sample mixtures of I and VI11 (possessing 0, 18, 50, 64 and 100% of V I I I ) that the ratio of the intensities of the bands at 6.1 and 6.25 w (designated as R ) was proportional to the percentage of VI11 in the boronic acid mixture up to 64y0 VIII. T h e yo VI11 values listed in Table I were obtained from the empirical relationship: % VI11 = 4311. The accuracy of these analyses was particularly good for low percentages of VIII. Since R was not very sensitive to minor fluctuations in thickness of the pellet or the concentration of the sample in the pellet even the poorer values in the range of 60% VI11 should be within several per cent. of the true value. At high percentages of VI11 the shape of the spectrum in the 12.8-13.4 p region was a more reliable index of the composition than the bands at 6.1 and 6.25 p . (21) H. R . Snyder, hl. S. Konecky and W. J. Lennarz, THIS JOURNAL, 80, 3611 (1958).
EVANSTON. ILL.
DEPARTMENT OF NEUROSURGERY, MASSACHUSETTS GENERAL HOSPITAL AND MEDICAL SCHOOL]
THE
HARVARD
Stability and Synthesis of Phenylboronic Acids1 BY -4. H. SOLOWAY RECEIVED NOVEMBER 25, 1958 The preparation of several substituted carboxyphenylboronic acids is described. The marked stability of the carbonboron bond in such compounds is noted. Three boron-containing compounds with potential carcinostatic properties have been synthesized.
Introduction.-In an investigation designed to prepare organoboron compounds for possible utilization in the treatment of brain tumors by neutroncapture irradiation2-7 several monosubstituted phenylboronic acids were screened8 in C3H mice with subcutaneous gliomas. The purpose of that study was to determine the structure and position of the substituted groups on the boronic acid which, in animal tests, resulted in a high tumor to brain differential boron concentration. An increased concentration in the neoplasm relative to the brain would permit selective destruction of tumor cells adjacent to normal nervous tissue upon thermal neutron irradiation. The feasibility of this procedure depends upon a difference in permeability between normal and neoplastic cells. In the case of brain and brain tumor, such a difference exists (1) This research was supported by
U. S. Atomic Energy Commis-
sion under contract h'o. A T (30-1)-1093 and by an Institutional Grant from the American Cancer Society.
(2) W.H. Sweet and M. Javid, Trans. A m . Neurol., 8 7 6 , 60 (1951). (3) W. H. Sweet and hl. Javid, J . Neurostirg., 9, 200 (1952). (4) P. A. Zahl, F. S. Cooper and J. R . Dunning, Puoc. Null. Acad. Sci. U .S . , 26, 589 (1940). ( 5 ) P. G. Kruger, ibid., 26, 181 (1940). (6) M . Javid, G. L. Brownell and W. H. Sweet, J . Clin. Invesl., 31, GO4 (1952). (7) L. E. Farr, W.H. Sweet, J. S. Robertson, C. G. Foster, R. B. Locksley, D. L. Sutherland, M. L. Mendelsohn and E. E. Stickley, A m . J . Roentgenol. Radium Therapy Nuclear M e d . , 71, 279 (1954). (8) A . H. Soloway, Science, 128, 1572 (1958).
because of an alteration in the blood-brain barrier in the n e o p l a ~ m . ~ JOf ~ the compounds tested,8 those with a carboxyl group in the m- or p-position or a meta-ureido function offered the most promise with regard to toxicity and tumor to brain boron ratio. Consequently, the synthesis of other carboxyphenylboronic acids was undertaken. Synthesis and Discussion.-The introduction of substituents into the nucleus or side chain of an aromatic boronic acid and the reactions which such compounds can undergo are very limited. The cleavage of the carbon-boron bond by a variety of agents such as acid, base, water a t elevated temperatures, hydrogen peroxide, halogens, salts of the B subgroup elements and organometallic compounds is a facile process and has been observed by many Attempts to prepare a boron-containing purine or pyrimidine by the interaction a t elevated temperatures of m-aminophenylboronic acid with 6-methylmercaptopurine (9) H. B. Locksley and W. H. Sweet, Proc. S O C .E x p . Biol. and Med., 86, 56 (1954). (10) G. E. Moore, Science, 106, 130 (1947). (11) M. F. Lappert, Chem. Revs., 66, 959 (1956). (12) A. D. Ainley and F. Challenger, J. Chem. SOC.,2171 (1930). (13) J. K. Johnson, M. G . van Campen and 0. Grummitt, THIS 60, 111 (1938). (14) L. Santucci and H. Gilman, ibid., 80, 193 (1958). (15) H. Gilman, D. R . Swayampati and R. 0. Ranck. ibid., 80, 1355 (195s).
JOURNAL,
and a150 with 2,4-clithiouracil 111 a riitrogeli ,It- ciirrespotiding aromatic amine occurretl readily tnosphere yielded only (i-anilinopurine anrl 2- without the rupture of the carbon--boron liiikagc. ~iiercapto-.l--aiiil~ii~~~~yriiiii~liiie~ rclspccti\ cl1 ' ~ I I I ~ liy tlie usual iiietliotl for forming aroiiiatic iircas,"' is yet another itist'incc ( ~ fhon 1.~1)ilc. the c,irhoii :I-ureidoglicii~ll~~Jr(liic. acid (IV) an(I ~~-iirc.icIo-/~boron linkage is. tolylboronic acid (V) were syiithesized; deboroThis instability is niarkcdly J t c ~ r e dby the 111- 11:itioii rcsultctl i n the formation of the corresporltltroduction of a carboxyl group i i i t o pheiiylboronic ing ureas. Tile proniising biological results8 of acid. Phenylboronic acid itself c m be nitratctl conipountl I T - led to the attempted preparation of a by fuming nitric acid at n temperature of - 10 to Ixiroiiic acid which would contain both a carboxyl - 13°,16 b u t a t higher temperatures extensive function and a ureido group. The synthetic decomposition occurs. y-CarbosypheiivlboriJnie schenic. which was uiidertaken was the catalytic acid, on the other hand, is readily nitrated J t rooiii reduction of .7-~1itro-l-carboxyphen).lboronic aci,] temperature by a mixture of fuming nitric acid .mtI ( 1 TI) to the kiiown18 ~~-amino-4-carboxypliei~y~sulfuric acids to yield a single 1)roduct Z-nitro--- I~oronicacid (VI) and its conversion to :3-ureitIo-4carboxyphenylboronic acid This compound I m 1 carbo.uypheriylboroiiic acid (VII). T h e amino acic] been previously preparedI7 aiitl its stability :ti well T"1 did react with potassium cyanate in an acetic as t h a t of ~-carboxyplieriylboronlcacid W ~ Sob- :iciil solution to forin a product. However, all served. The attempted nitr,itioii of nz-nitrophenyl- attempts a t the crystallization of this phenylurc:~ boronic acid and phenylbormic acid, itself, by this were without success; oiily a gel could be isolated. procedure resulted in the foriiiJtion ( J f ?~?-dinitro- Preparation of :tii antimetabolite or any carbeiiyeiie as the only isola1)lc. product 111 similar cinostatic agent contaiiiing boron may permit this tic:utroii-capture therapy to be applied to neonianner, ? ~ z - c a r b o x y p h c i i ~ ~ ~acid ~ ~ ~w,is r o l ~nitrated ic other than of the central nervous systeiii. to yield the J\iiowii1' 2 iiitrri 5 r ~ r l ~ r ~ x y p l i c ~ i plasiiis yltwcdold attack 011 the tumor might be fcasiblc ; boronic acid ( t I ) . Xiiother instance of the liicreascd st~ibility(Jf tlic mie, iiihibitioii o f growth: ancl two, incorporation ( t f tlic aiitiiiic.tabc11itc."' or c:trcinostntic agent i i i t c i carbon-boron linkage in carbouyl-sti1,stltutetl phcn ylboronic acids was o b s m cd 111 attciiipti ,it t!ic nuclei or iicop1:istic cells. Siicli a nuclcnr p s i :I lxroii''' atoiii by the inctnbolisni ( J f tlic chlorosulfonation Of aroinatic borunic acids. .it tioii O", p-tolylboronic acid reacted with chlorcisulfoiiic tuiiior woultl allow chroiiiosonc~ disruption ripon acid to form p-toluenesulfonyl chloride. The struc- thcriiial neiitroii irratliatiori aiitl may ~iossi1,lyperture of the product was verified by its conversicm mit l o w r ncutrori fluxcs for cradication of thv to p-toluenesulfoiianide. Under identical reaction neoplnsnl. On this basis, a s y i thctic scheme directed toward conditions, p-carboxyplienylborcJtiic acid was recovered unchanged. Consequen tly, there woiilti tlic preparation of a purine antiiiietabolite conappear to be a marked difference in the stability of taining boron was ~ i i d e r t a k e n . ? ~The recent work the carbon-boron bond in aromatic boronic acids. of H. R . Snyder, et a l . , 2 4with the synthesis 0.1 Both 4-dimethylamino- and 4nicthoxy- a-naphthyl- 1,oroiioI)henylalanine presents a very interesting apboronic acids were easily c!eboronatedlq by inild proach to this same probleiii. -i--B~)ronoanthranilic acid or alkaline hydrolysis. Likewise, the bromiii- acid (VI) may also be an ainino acid antagonist inolysis of substituted phenylboronic acids shows thcit hibiting utilization of tryptophan. Other metabolic antagonists and carcinostatic those compounds with electron-releasing qroups have appreciably greater rate constants The agents were also considered. N-Phenylurethan has increased stability of p-carboxypiieiiylboronicacid s h o ~ v i iantimitotic action itself?: and the urethans to electrophilic displacernent of the bcroii moiety as a class of compounds have been eniployed in is understandable on this basis. Howex er. its clinical trials?6 in the treatment of cancer. 3greater stability in sulfuric -nitric acid niixtures Carl~ethoxyarninoplienylb(~ronicacid (VI I I), which relative to m-nitro1)hcnylborotiic acid requires a had been prepared,?: was synthesized directly from acid. consideration of other factors, possibly tlie ~ 1 ~ ~~n-aiiiinn~~lienylboronic 1 In l-iew of the incorporation of nicotinamide antion of the carboxyl function. A third tiitrocarboxyphenyIbororiiic acid isomer tagonists into neoplastic cells??the synthesis of such which had been describedIs was prepared by thc a conipound containing boron was undertaken. oxidation of 3-tiitro-p-tolylht,roiic acid with potas- For this purpose, nz-aminophenq.lboronic acid was sium permanganate to :~-iiitro-~-carhoxyI,henyl-acylated with nicotinoyl chloride to yield 3boronic acid (111). The perinangariate oxidation nicotinaniidophenyJbor[~riic ncid (IX). T h e carof the methyl group of p-tolylboronic acid was a cinostatic activity of this compound, 4-boronomuch more facile process than t h a t with a nitro anthranilic acid :tnd the boron-containing urethan substituent ortho to the methyl group. The latter hnre not been evaliiated as yet. : 2 1 ) A . I Vu-el, ".4 Textho,,k of Proctical Organic Chemistry," oxidation required longer reaction times and higher 1,ongmans Green a n d C o , Xcm York, N . I'.. l W 8 , pp. (i16--618. temperatures to forin the correspontlinq aromatic i?) N. 0. Kap!an, A . G d d i n -i. K . H u m p h r e y s , A I . X I . C i o l l i :ind carboxylic acid. J . AI. \'enditti, S c i e n c r . 120, 437 ( 1 9 . j i ) . !L'.%) R m r r y K ) i l n s ;rnd A I € St21tn\ay,TtrIs J O U K X A I . , 81, 2081 The preparation of a ureido group froni thc f10) W Seaman a n d J
R Johnson ' h i s J o r
h i 11
( 1 7 ) K Torsiell A ? k i K e m t 10 1 7 8 i l O i 7 ) 118) K 'Ior350°, and m-carbethoxi-to Dr. IViIliani H . Sweet, Associate Professor in Sur-
O& r.7
I
~
(28) tube. (29) (30) (31) (32) (33) (1934).
Melting points are uncorrected and were obtained by capillary F. R. Bean and J. R. Johnson, THIS J O U R N A L , 6 4 , 4415 (1932). M. F. Lappert, C h e m . Revs., 56, 959 (1956). 63, 711 (1931). A . Michaelis, A n n . , 915, 38 (1901). B. Bettman and G . E. K Branch, THISJ O I J R N A I . , 56, 18GA
w. Seaman a n d J. R . Johnson, THISJ O U R N A L ,
gery a t the Harvard Medical School, for his great interest and encouragement. The technical assistance of Mr. Paul Szabady was most helpful. BOSTOS14, MASS (34) One gram of silver nitrate was dissolved in 8 ml. of n a t e r and this m a s clilritetl t o 1 0 m l . with 2 x 7 aqiie(iiis ammoniu.
